Cancer & Causality

A review of Causality in Medicine with Particular Reference to the Viral Causation of Cancers, by Brendan Clarke.

Brendan Clarke tackles one of the main and most controversial issues both in general philosophy of science and in the philosophy of medicine, that is causation. His remarkably clear and detailed analysis takes cancer research as a privileged standpoint, focusing on how we are to conceive the modeling of a causal nexus when dealing with the viral etiology of cancer. Reflections in philosophy of science are here substantiated by specific examples, described with great historical and medical precision. Clarke’s doctoral thesis is strongly interdisciplinary, successfully integrating history and philosophy of science. By constantly intertwining considerations stemming from the two fields, flanked by strong medical competence, a view on medical causation is built up through the chapters. Given the breadth of the current philosophical debate on causation, Clarke starts off by clarifying his main concerns: his investigations on the nature and role of evidence in the assessment of causal relations are carried out with an eye to the operational features of causes and how we can most effectively exploit causal knowledge in medicine.

Ed. Note: At the request of the author, Brendan Clarke, we are happy to include a link to the full version of the dissertation as well.

Before addressing the strictly philosophical literature and perspectives on the topic, two chapters (2 and 3) are devoted to extensive reconstructions of the building of mechanistic models for two cancers, Burkitt’s lymphoma and cervical cancer. In both cases, the role of viruses – namely Epstein-Barr and HSV – are examined by tracing the story of how they came to be part of the causal pictures accounting for the diseases. Clarke distinguishes various phases in causal discovery: suspicion, when researchers suspect that a new type of cause is at play; domain finding, when they detect the kinds of things over which the cause operates; making mechanisms, i.e. constructing a mechanistic model, which is finally applied to disease populations. By exploring different kinds of evidence – from, e.g., histology, epidemiology, clinical practice, … – its collection, uses and synthesis in the mechanistic models that were finally elaborated, Clarke comes to a critical examination – in Chapter 4 – of the main philosophical reference point of the thesis, namely the position put forward by Federica Russo and Jon Williamson (“Interpreting Causality in the Health Sciences,” International Studies in the Philosophy of Science 21 (2007), pp.157-170). They claim that “evidence (for causation) is constituted by two complementary elements: probabilities and mechanisms” (Russo and Williamson 2007, p.159), and that traditional causal monism does not accommodate such evidential pluralism in causal discovery. While largely agreeing with Russo and Williamson, and believing that their mechanistic/probabilistic distinction is tenable, Clarke presents us with various objections and counterexamples, evaluating their view with respect to actual cases of medical causation (McArdle disease; causal role of chronic HBV in primary liver cancer; causal role of HSV in cervical cancer).

The focus is then narrowed down (Chapter 5) to the mechanistic approach to causation and its appeal for medicine. Presenting and discussing mechanistic and neo-mechanistic philosophical views, and some suggestions from Jim Woodward’s interventionist position, Clarke tackles a number of features of these theories that can prove especially relevant for such complex phenomena as viral oncogeneses, emphasizing multilevel models, gappy descriptions, functional definitions, branching and track-switching, preventions and omissions. This leads Clarke to define mechanistic models as epistemic devices allowing us to conceptualize and organize complex causal relationships, deemed compatible with Russo and Williamson’s position. Finally, strongly endorsing evidential pluralism and following Phyllis McKay Illari (“Mechanistic Evidence: Disambiguating the Russo-Williamson Thesis,” International Studies in the Philosophy of Science 25 (2011), pp.139-157), Clarke distinguishes between different types of evidence – mechanistic evidence, evidence of mechanism, difference-making evidence, evidence of difference-making – and sheds light on how they combine in a web, and, by so doing, contribute to the assessment of medical causation and the construction of medical models. Faithful to his medical-historical standpoint, Clarke argues convincingly that “the sorts of evidence that are important for causation change chronologically, and that these chronological changes can be characterized in terms of the precise aspects of causal models that are being developed at the time” (p.151), and re-examines his case-studies in the light of the above mentioned notions of evidence. Russo and Williamson are surely the authors Clarke is directly responding to, by partly sharing, partly refining their view. It is thus concluded that the Russo-Williamson perspective can be historically motivated, given also a normative interpretation and taken to provide an adequate causal road-map, since it maintains that “causal claims that are supported by appropriate evidence of both mechanism and difference-making are very likely to be more reliable that those that are not” and might therefore make clear “in which circumstances a particular causal claim is likely to be an effective guide to practice, and when it is not” (p.183).

While Clarke makes specific and extensive reference to Federica Russo and Jon Williamson’s work, the background to the thesis and the framework it traces are much wider. Tackling such topics as medical causation and the relationship between evidence and mechanistic knowledge involves dealing with a number of problematic issues. Clarke provides insightful – historically supported – considerations on, for instance, discovery, explanation, prediction, description, classification, their place in medicine and the relationships holding between them. These matters also direct attention to the (increasingly recognized) role of the context, the relationships between observational and experimental evidence, and the relevance of the type/token distinction for medical causation, which all find room in Clarke’s analysis. The thesis delivers more than it expressly promises, by stimulating further discussion on other issues such as current accounts of building mechanistic models, the reductionism/anti-reductionism debate in studies on cancer causation, the use of simulations and animal models in assessing causal links. Rooted in philosophy of science and going hand-in-hand with history of science, Clarke’s “mapping [of] the causal territory” (p.138) has the merit both of giving a metascientific, non-reductivist account of causality in medicine, and of pointing to further challenges philosophy of medicine is called to face, highlighting the specificity of issues arising in medicine and, hence, the specificity of philosophy of medicine as a field of enquiry.

Raffaella Campaner
Department of Philosophy
University of Bologna

Primary Sources

Bibliographical analysis of peered-reviewed journal articles, monographs, edited volumes, conference proceedings.

Dissertation Information

University College London. 2010. 228pp. Primary Advisor: Donald Gillies.